An Adaptive Statistical Non-uniform Quantizer for Detail Wavelet Components in Lossy JPEG2000 Image Compression

The paper presents a non-linear quantization method for detail components in the JPEG2000 standard. The quantization step sizes are determined by actual statistics of the wavelet coefficients. Mean and standard deviation are the two statistical parameters used to obtain the step sizes. Moreover, weighted mean of the coefficients lying within the step size is chosen as quantized value contrary to the uniform quantizer. The empirical results are compared using mean squared error of uniformly quantized and non-linearly quantized wavelet coefficients w.r.t. original wavelet coefficients. Through empirical results, it has been concluded that for low bit rate compression, the quantization error introduced by uniform quantizer is higher than that of non-linear quantizer, and thus suggesting the use of non-linear quantizer for low bit rate lossy compression.